Neostriatal GABAergic Interneurons Mediate Cholinergic Inhibition of Spiny Projection Neurons

Abstract

Synchronous optogenetic activation of striatal cholinergic interneurons ex vivo produces a disynaptic inhibition of spiny projection neurons composed of biophysically distinct GABAAfast and GABAAslow components. This has been shown to be due, at least in part, to activation of nicotinic receptors on GABAergic NPY-neurogliaform interneurons that monosynaptically inhibit striatal spiny projection neurons. Recently, it has been proposed that a significant proportion of this inhibition is actually mediated by activation of presynaptic nicotinic receptors on nigrostriatal terminals that evoke GABA release from the terminals of the dopaminergic nigrostriatal pathway. To disambiguate these the two mechanisms, we crossed mice in which channelrhodopsin is endogenously expressed in cholinergic neurons with Htr3a-Cre mice, in which Cre is selectively targeted to several populations of striatal GABAergic interneurons, including the striatal NPY-neurogliaform interneuron. Htr3a-Cre mice were then virally transduced to express halorhodopsin to allow activation of channelrhodopsin and halorhodopsin, individually or simultaneously. Thus we were able to optogenetically disconnect the interneuron-spiny projection neuron (SPN) cell circuit on a trial-by-trial basis. As expected, optogenetic activation of cholinergic interneurons produced inhibitory currents in SPNs. During simultaneous inhibition of GABAergic interneurons with halorhodopsin, we observed a large, sometimes near complete reduction in both fast and slow components of the cholinergic-evoked inhibition, and a delay in IPSC latency. This demonstrates that the majority of cholinergic-evoked striatal GABAergic inhibition is derived from GABAergic interneurons. These results also reinforce the notion that a semiautonomous circuit of striatal GABAergic interneurons is responsible for transmitting behaviorally relevant cholinergic signals to spiny projection neurons.

Significance statement: The circuitry between neurons of the striatum has been recently described to be far more complex than originally imagined. One example of this phenomenon is that striatal cholinergic interneurons have been shown to provide intrinsic nicotinic excitation of local GABAergic interneurons, which then inhibit the projection neurons of the striatum. As deficits of cholinergic interneurons are reported in patients with Tourette syndrome, the normal functions of these interneurons are of great interest. Whether this novel route of nicotinic input constitutes a major output of cholinergic interneurons remains unknown. The study addressed this question using excitatory and inhibitory optogenetic technology, so that cholinergic interneurons could be selectively activated and GABAergic interneurons selectively inhibited to determine the causal relationship in this circuit.

Keywords: GABA-Aslow; cholinergic interneuron; neurogliaform; optogenetics; striatum.

Figure 1.

Schematic for examining interneuronal contributions to cholinergic-induced inhibition. A, Diagram of tissue from double-transgenic animals in which cholinergic interneurons (CHAT, dark blue) are activated with a focal blue LED and Htr3a-Cre interneurons (green) are inhibited with a wide angle yellow LED (yellow). B, Example of decoupling an interneuron from the nicotinic circuit. Top, blue bar: 2 ms blue LED stimulus; top, yellow bar: 1 s yellow LED stimulus. Activation of CINs produces nicotinic EPSP and action potentials in the fast-adapting interneuron (black traces). Concurrent inhibition with yellow light hyperpolarizes neuron and prevents EPSP from reaching spike threshold (yellow traces). Inset, Current–voltage responses of the HR3.0-expressing fast-adapting interneuron.

Figure 2.

Cholinergic-induced inhibition in SPNs before and after inhibition of Htr3a-Cre interneurons. A, Voltage-clamp recordings of SPNs from Dbl-Tgn mice following cholinergic activation. Top, blue bar: 1–2 ms blue LED stimulus; top, yellow bar: 1 s yellow LED stimulus. Thin gray traces represent individual control trials following blue LED pulse. Thin yellow traces represent individual HR3.0 trials during simultaneous blue and yellow LED pulses. These two trial types were acquired on an alternating schedule. Black traces represent average of control trials (n = 9, excluding first trial). Red traces represent average of HR3.0 trials (n = 10). B, Voltage-clamp recording of SPN from single-transgenic ChAT-ChR2 mouse following cholinergic activation, using the same color scheme used as in A. Note the lack of IPSC reduction during yellow LED pulses.

Figure 3.

Inhibition of Htr3a-Cre interneurons delays cholinergic-induced IPSC. A, Example SPN from Dbl-Tgn mouse, using same color scheme as in Figure 2. Note that the main effect of HR3.0 in this neuron was the reduction of the fast component and the delay in its onset and peak, observable in the inset. B, Example SPN from Dbl-Tgn mouse, using same color scheme as in Figure 2. Note the increased delay of IPSC onset (black arrow) and peak (white arrows) primarily on HR3.0 trials. Note also the failures experienced by this SPN on both trial types, indicative of unitary sources (individual neurons) failing to fire action potentials.

Figure 4.

Statistical analysis of HR3.0 effect on IPSC kinetics. A, Example voltage-clamp recording of SPN from Dbl-Tgn mouse as in Figure 2, following the same color scheme. For this SPN, note the presence of both GABA_A fast and GABA_A slow peaks for the control traces (black, gray). The IPSC fast-peak amplitude, latency, and slow-peak interval used to quantify IPSC amplitude and delay are detailed. Blue bar: 2 ms blue LED stimulus. B–D, Left, Within-group examination of HR3.0 effect on IPSC amplitude and latency for individual Dbl-Tgn SPNs. Black traces represent average responses in SPNs, where HR3.0 significantly affected the IPSC amplitude or latency (p < 0.05, paired-sample t test or paired-sample Wilcoxon signed rank test of individual control and HR3.0 trials, such as the thin gray and yellow traces in A). Red traces represent average responses in SPNs, where HR3.0 failed to do so (p > 0.05). **p < 0.01; paired-sample Wilcoxon signed rank test, using average responses for each SPN. Right, Between-group examination of normalized percentage IPSC reduction or delay in SPNs from Dbl-Tgn mice or single-transgenic ChAT-ChR2 control mice. Box plot bars represent minimum (bottom attached bar), Q1 (box bottom), Q2 (middle line), Q3 (box top), and maximum values (top attached bar). Mean represented by central dot. Detached bars represent outliers (values lesser or greater than Q1, Q3 ± 1.5 IQR). Note the large range in slow- and fast-peak IPSC reduction for the Dbl-Tgn group. **p < 0.01, Mann–Whitney test; *p < 0.05, two sample t test.

Figure 5.

Relationship between IPSC reduction and amplitude. A, Correlation between the IPSC first acquired trace (largest gray traces; Fig. 2A) and percentage reduction for the fast peak. B, Correlation between the IPSC first acquired trace (largest gray traces; Fig. 2A) and percentage reduction for the slow peak. Although there are negative trends in both A and B, no significant correlations were observed. C, Correlation between fast-peak and slow-peak percentage reduction for all SPNs exhibiting significant reductions in at least one of these two components. Note the strong linear correlation between these two responses.